| Nowadays, Intelligent Transportation System (ITS) is one of the most important traffic management and control strategies. With the dissemination of mass information, road users can make their decisions to save their travel time or plan their trips. How to provide real time and precise prediction information to traveler becomes an important issue in ITS.; Traffic speed impact caused by adverse weather is the main subject of this study. In order to understand the impact according to various types of weather condition, an artificial neural network (ANN) model for predicting traffic speed with weather conditions as input is proposed. According to the analysis results of numerical experiments, an ANN model has served this case well.; In this study, five links were selected for analysis, which located in Chicago, Seattle and Minneapolis. All the data for these links, such as weather condition and traffic speed, was collected from the Internet. Several JAVA programs were developed for this purpose.; The Back Propagation Algorithm is used in this ANN model to adjust the weight in order to achieve the global minimum of the error surface. This means the model can be used to predict the traffic speed with minimum error. The MATLAB software was used to implement this ANN model.; The results have demonstrated that the ANN model can precisely estimate the traffic speed based on historic data and can help easily figuring out the traffic impact. In addition, the results indicate that snow condition will have the largest impact on traffic speed. |
